Principles and Practice of Pharmaceutical Medicine

are made in discussions among committee members
and may not necessarily be based on hard and fast
criteria. Also, unlike a computer flow chart, the
decision concerning a particular drug will usually
be based in part on the results of work with other
compounds that have the same indication.
In the boxes representing tasks to complete in
the phase I study in humans, we have used the
symbol 1 to represent work that can be expedited
by good validated preclinical data. The symbol 2
represents the tasks that can be expedited by online
pharmacokinetic modeling. Among the pharmaco-
kinetic questions that will be asked online in the
phase I trial are the following:

1. As the doses are escalated, do the kinetics of the
   drug appear to be linear or nonlinear over the
   dose range?


2. With repeated dosing, is there any evidence of a
   change in kinetics, for example a higher elim-
   ination rate that might be indicative of autoin-
   duction?


3. Does the drug accumulate in tissues more than
   predicted with repeated dosing?
   4. If preclinical work identified metabolite(s) to
   measure in humans, are the pharmacokinetics
   of metabolite(s) linear and as predicted?
   5. Does the relationship between concentration
   and effect change with dose, time and duration
   of treatment?

We expect that the task lists represented by some

of the boxes will increase. For example, within the

box including ‘in vitrointrinsic clearance’, there

may bein vitropredictors of oral availability and

measures of potentially toxic metabolites. The ‘in

vivopharmacokinetics’ in rats may include an

increasing number of compartments whose con-

centrations are measured by microdialysis and may

include measures of a few selected metabolite

concentrations.

This diagram isnota comprehensive guide to

drug discovery. However, it does show that the

chemists discover new chemical entities with

desirable properties.In vitrobiochemistry is fol-

lowed by initialin vivowork in the rat which is

conducted with pharmacokinetic support andin

vitro drug metabolism in parallel. Compounds

meeting pre-arranged criteria proceed through

1000

100

10

1

0.5

0.5 1 10 100 1000

B

1A

M

F

O

C

ki (ng/ml)

EC

50

u (ng/ml)

Figure 8.7 Correlation (r¼ 0 :993,p< 0 :001) between benzodiazepine-free drug concentrations EC 50 units produ-
cing 50% of the maximal EEG effect (change in amplitudes in the aˆfrequency band, as determined by aperiodic EEG
analysis) and affinity to the GABA–benzodiazepine receptor complex (Ki). Binding to the benzodiazepine receptor was
determined on basis of displacement of [^3 H]flumazenil in washed brain homogenate at 37C for six drugs B, IA, M, F, O,
and C. (Reproduced with permission from Danhof and Mandema, 1995)

96 CH8 PHASE I: THE FIRST OPPORTUNITY FOR EXTRAPOLATION FROM ANIMAL DATA